Gold nanoring-enhanced generation of singlet oxygen: an intricate correlation with surface plasmon resonance and polyelectrolyte bilayers

Abstract

We report the dependence of gold nanoring (Au NR)-enhanced singlet oxygen (¹O₂) generation on localized surface plasmon resonance (LSPR) wavelength and separation distance between Au NR and photosensitizers. Using Au NR of tunable LSPR and poly(styrene sulfonate) (PSS) and poly(allylamine hydrochloride) (PAH) bilayers via layer-by-layer (LbL) assembly on Au NR as molecular spacers, we found that ¹O₂ generation from Al(III) phthalocyanine chloride tetrasulfonic acid (AlPcS₄) conjugated on the Au NR can be optimized through proper combination of LSPR and the separation distance. More importantly, ¹O₂ enhancement follows a different correlation with the separation distance compared to that of AlPcS₄ fluorescence enhancement. Our experimental findings are consistent with the prediction by boundary element method (BEM) calculations. Our study paves the way to the design of LSPR-enhanced ¹O₂ generation for improved efficacy of photodynamic therapy of cancers.

---

Introduction

Singlet oxygen (¹O₂) plays an indispensable role in numerous photochemical processes in biological systems due to its strong oxidative nature.^1–5 Reliance on ¹O₂ in photodynamic therapy (PDT) for cancer treatment is a good example. The production of ¹O₂ perturbs cellular processes and leads to the destruction of cells via apoptosis or necrosis.^6,7 There is a growing interest in the scientific and medical research communities in exploring and exploiting means of enhanced production of ¹O₂ for PDT of cancers, amongst other applications.

It has been shown that plasmonic nanoparticles promote elevated generation of ¹O₂ from photosensitizers (PS) upon light irradiation.⁸ Our previous findings have also indicated that ¹O₂ generation can be significantly increased by incorporating gold nanostructures such as gold nanoparticle^9–12 and gold nanorings (Au NR),¹³ due to localized electromagnetic (EM) field enhancement. We have also revealed that the photosensitivity of PS and ¹O₂ generation can be inhibited in PS–Au NR conjugates when there is a substantial overlap between the emission wavelength of PS and localized surface plasmon resonance (LSPR) of Au NR, leading to the radiative and non-radiative energy transfer from excited PS to Au NR.¹⁴ Once the PS is released from the surface of Au NR, ¹O₂ generation of PS will be recovered and can be further enhanced at a finite distance from Au NR. Similar phenomena have been reported by Jang and co-workers in PS–gold nanorod system.¹⁵ Clearly, enhancement in ¹O₂ generation by plasmonic nanostructure depends on both the LSPR of the nanostructure and the distance between PS and the nanostructure itself. There is a dearth of systematic study to provide insights into this significant correlation.

Here we present an integrated experimental and numerical investigation on the effect of PS-nanostructure separation distance and LSPR on ¹O₂ generation using polyelectrolyte multilayers as spacer and Au NR as EM enhancer. The distance between PS and Au NR is adjusted by layer-by-layer (LbL) assembly with a high degree of control due to the nature of interaction between oppositely charged polyelectrolytes. This approach allows us to measure the spatial extent of PS–Au NR interactions at the nanoscale and its correlation to ¹O₂ generation. The LSPR of Au NR can be easily tuned over a wide range, from the visible to near-infrared (NIR) wavelength, by adjusting their size and aspect ratio.¹⁶ This freedom enables us to manipulate the radiative and non-radiative energy transfer between PS and Au NR. In addition, boundary element method (BEM) was employed to establish the relationship between the elevated generation of ¹O₂ and the PS molecule transition rates modified by coupling with a nearby Au NR, indicating different and competitive energy transition pathway between Au NR-enhanced ¹O₂ generation and Au NR-enhanced PS fluorescence.

Experimental

Materials

Cobalt chloride hexahydrate (CoCl₂·6H₂O, 99.99%), sodium borohydride (NaBH₄, 99%), gold(III) chloride solution (30 wt% of HAuCl₄ in diluted HCl), poly(vinylpyrrolidone) (PVP, M_w = 2500), poly(allylamine hydrochloride) (PAH, M_w ∼ 15 000), and poly(sodium 4-styrenesulfonate) (PSS, M_w ∼ 70 000) were purchased from Sigma-Aldrich (St. Louis, MO). Sodium citrate trihydrate was procured from Fisher Scientific (Pittsburgh, PA). Al(III) phthalocyanine chloride tetrasulfonic acid (AlPcS₄) was obtained from Frontier Scientific Inc (Logan, UT). Singlet oxygen sensor green (SOSG) was purchased from Life Technologies (Grand Island, NY). Silica Microspheres (SiO₂, size ∼ 50 nm) were purchased from Polysciences, Inc. (Warrington, PA). Milli-Q ultrapure water (less than 18.2 MΩ cm) was used for synthesis experiments.

Preparation of Au NR

Au NR was prepared using galvanic replacement reaction as described previously.¹⁶ Briefly, a mixture of 100 μL 0.4 M CoCl₂·6H₂O and 400 μL 0.1 M trisodium citrate dihydrate was added to 100 mL Milli-Q water and deaerated for 40 min. Then, 1 mL of 0.1 M freshly prepared NaBH₄ and 200 μL of 1 wt% PVP were injected into the solution simultaneously under vigorous mechanical stirring with continuous argon flow to prevent oxidation, resulting in the formation of Co nanoparticles. After stirring for 40 min, 150 μL of 0.1 M HAuCl₄ was added into the cobalt solution drop wise and reacted for 30 min. Thereafter, the solution was exposed to the ambient environment to remove unreacted cobalt by oxidation and dissolution of cobalt oxide, thereby obtaining highly pure colloidal Au NR.

Characterization of Au NR

The size, distribution and absorption spectra of as-fabricated Au NR were examined with a transmission electron microscope (TEM, Philips CM20) and a UV-vis spectrometer (Synergy HT multidetection microplate reader, BioTek Instruments, Inc., Winooski, VT). The obtained images were analyzed for size and distribution with Image-J 1.46 software (NIH).

Au NR immobilization and LbL of polyelectrolyte multilayer spacers

¹O₂ oxygen generation measurements were carried out using PS conjugated with Au NR with or without polyelectrolyte spacers immobilized in 96-well plates. Each well from the 96-well plate was first modified with a monolayer of positively charged PAH. PAH was dissolved in 0.1 M NaCl at a concentration of 0.2 mg mL⁻¹. The pH of PAH was adjusted to 7. After 20 min of deposition, the substrate was washed three times with Milli-Q water followed by deposition of monolayer Au NR. Immobilization was achieved by electrostatic attraction between PAH and slightly negatively charged Au NR (ζ-potential of −5 mV in Milli-Q water) as well as by the binding affinity of Au to amino groups in PAH.

For polyelectrolyte deposition on immobilized Au NR in the well plate, PSS and PAH were dissolved in 0.1 M NaCl solution. Concentrations of both solutions were 0.2 mg mL⁻¹ with pH ∼ 6.5. LbL of polyelectrolyte thin film spacers began with depositing PAH as anchor layer from its solution for 20 min, followed by alternate deposition of PSS/PAH bilayers. The deposition for each layer was 10 min followed by thorough rinsing in Milli-Q water. The deposition was terminated with a PAH layer.

Polyelectrolyte film thickness by ellipsometry

Exactly the same procedures for the deposition of PSS/PAH bilayers on Au NR as described above were followed for their deposition on Si substrate to allow ellipsometry measurements of the thin film, or spacer, thickness. Si substrates were cleaned as reported previously.¹⁷ The deposition process was repeated until the desired number of bilayers was achieved.

The measurement of LbL film thickness was performed using a custom-built, single-wavelength, phase-modulated ellipsometer at 65° angle of incidence. The refractive indices for the native oxide layer on Si and the polyelectrolyte polymer films were 1.333 and 1.500, respectively.

¹O₂ generation and fluorescence measurements

The Au NR immobilized inside the well plate and subsequently coated with PAA/PSS bilayers and AlPcS₄ photosensitizer were utilized for the quantitative measurements of ¹O₂ generation. SOSG was used as a global ¹O₂ tracking agent. Briefly, 50 μL of 10 μM AlPcS₄ was added into 96-well plates (n = 3) and allowed for its surface adsorption for 30 min. The fluorescence intensity of AlPcS₄ was first taken using Synergy™ multi-mode microplate reader at 660/10 nm for excitation and 690/10 nm for emission. Then, 50 μL SOSG was added to the samples at a final concentration 1.33 μM. All the samples were irradiated for different time durations with a 150 W halogen lamp (100 mW cm⁻², Dolan-Jenner Fiber-Lite MI-150, Dolan-Jenner Industries, MA) filtered through 600 nm long pass filter (Thorlab FEL0600). The fluorescence intensity of SOSG was measured after irradiation using the Synergy™ multi-mode microplate reader at 485/20 nm for excitation and 528/20 nm for emission.

Statistical analysis

Each experiment was repeated at least 3 times and data were expressed as the mean ± standard deviation. All quantitative measurements were collected at least in triplicate for each group.

Boundary element method (BEM) simulation

The BEM method was used to calculate the optical properties of Au NR (cross sections and near-field enhancement) and AlPcS₄–Au NR system (radiative decay rate). The BEM was implemented as a Matlab toolbox.¹⁸ This technique is convenient to solve Maxwell's equations for a dielectric environment where bodies with homogeneous and isotropic dielectric functions are separated by abrupt interfaces. The radiative and non-radiative decay rates of AlPcS₄, modified by coupling with Au NR, were calculated using the dyadic Green function approach.

The lifetime (τ⁰_Fluo) and quantum yield (Q⁰_Fluo) of AlPcS₄ in aqueous medium are 5.2 ns and 0.5, respectively, according to the literature.^19,20 The dielectric function of Au NR is taken from measurements by Johnson and Christy.²¹

Results and discussions

Experimental

In this work, we took advantage of the well-established LbL method to fabricate the samples with various controlled separations between PS and the Au NR surface, enabling the study of the distance effect on Au NR-enhanced ¹O₂ generation. Multilayered structures of poly(styrene sulfonate) (PSS) and poly(allylamine hydrochloride) (PAH) were fabricated with molecular precision. In order to avoid the instability and poor monodispersity after repeated purification processes during LbL on freestanding colloidal nanoparticles, we anchored individual Au NR on planar substrate. The distance between PS and the Au NR surface was modulated by varying the number of PSS/PAH bilayers (see Fig. 1). Before deposition onto the Au NR, we first examined the thickness of PSS/PAH multilayers on Si with a native oxide. Multilayer assemblies consisting of up to 20 bilayers were prepared by consecutive adsorption at neutral pH (∼6.5) and measured by in situ ellipsometry. The thickness of PSS/PAH films followed a linear dependence on the number of deposited bilayers (Fig. 2a and Table S1, ESI†), in agreement with literature reports.^22,23 Each PSS/PAH bilayer contributes to a thickness of ∼2 nm. Owing to the dramatic distance-dependent decay of EM field intensity of plasmonic nanoparticles at far field (>∼10 nm),^24,25 we focused on 1 to 5 bilayers for investigation of Au NR-enhanced ¹O₂ generation.

Fig. 1 Schematic representation of sample preparation for distance-dependent Au NR-enhanced ¹O₂ generation using PSS/PAH multilayers as nanoscale spacer.

Fig. 2 (a) Film thickness of PSS/PAH in H₂O as measured by in situ ellipsometry; (b) absorption–emission spectra of AlPcS₄ (λ_max-Ab = 675 nm and λ_max-Em = 685 nm).

Construction of LbL layers of PSS/PAH on Au NR immobilized on the substrate and subsequent PS absorption were performed conveniently as shown in Fig. 1. Al(III) phthalocyanine chloride tetrasulfonic acid (AlPcS₄), a second-generation PS with a high ¹O₂ yield, was chosen. Fig. 2b shows typical absorption and emission band maxima of AlPcS₄ at 675 nm and 685 nm, respectively. Shown in Table S2 (see ESI†) and Fig. 3 are a set of Au NR we synthesized with varying sizes and outer/inner diameter aspect ratios, due to the dimension effect on the LSPR behavior in published studies.^16,26–28 As the aspect ratio increases from (Au NR I) to (Au NR V), the LSPR wavelength of Au NR is largely blue-shifted from 823 nm to 574 nm. When the aspect ratio is similar, for example, Au NR II vs. Au NR III, and Au NR IV vs. Au NR V, the decrease in the NR height would lead to the blue-shift of Au NR LSPR wavelength as well. The LSPR variation covers the absorption and emission range of AlPcS₄ (Fig. 2b). Before light irradiation, we investigated the fluorescence enhancement of AlPcS₄ first. Shown in Fig. 3 is the fluorescence intensity of AlPcS₄ bound to Au NR at various distances from Au NR surface, as dictated by the PSS/PAH bilayers. For all the measurements shown in Fig. 3, the fluorescence intensities of AlPcS₄ in the present of Au NR is distance-dependent, consistent with literature reports.^24,29 Interestingly, the highest fluorescence intensity of AlPcS₄ from each group does not follow the same trend. For Au NR I (Fig. 3a) with LSPR peak at 823 nm, the highest fluorescence enhancement occurred when AlPcS₄ was positioned at 2 bilayers (∼3 nm). Similar results were observed from Au NR IV (Fig. 3d) and Au NR V (Fig. 3e), whereas the same magnitudes of fluorescence enhancements were observed at 3 bilayers (∼5.4 nm) for Au NR II (Fig. 3b) and Au NR III (Fig. 3c). Upon reaching fluorescence maximum of AlPcS₄, its intensity decreased with an increase in the number of PSS/PAH bilayers. A monotonic reduction in the fluorescence intensity also followed when AlPcS₄ was in close proximity to the Au NR in all the measurements. This could be related to the quenching of emission when the fluorophores get very close to metal surface, as a result of non-radiative resonance energy transfer (RET) due to dipole–dipole interaction.³⁰

Fig. 3 Dependence of ¹O₂ generation on Au NR nanostructure: (a) Au NR I, (b) Au NR II, (c) Au NR III, (d) Au NR IV, and (e) Au NR V. (f) 50 nm SiO₂ nanoparticles as control. From left to right, 1st column: TEM images; 2nd column: absorption spectra; 3rd column: fluorescence intensity of AlPcS₄ when conjugated with Au NR coated with the polyelectrolyte thin film spacers before the light irradiation; 4th column: fluorescence intensity of SOSG indicating ¹O₂ generation after 10 min light (λ > 600 nm) irradiation.

Singlet oxygen sensor green (SOSG), a global ¹O₂ tracking probe,³¹ was applied to indirectly quantify ¹O₂ generation. After exposure to near-infrared (NIR) light (λ > 600 nm) for 10 min, the fluorescence emission intensity of SOSG was measured and correlated with ¹O₂ generation. Surprisingly, dependence of ¹O₂ generation and AlPcS₄ fluorescence intensity on separation distance between AlPcS₄ and Au NR in the same group did not follow the same trend as shown in Fig. 3. In other words, the distance for the largest enhancement for ¹O₂ yield was different from the one for the highest fluorescence intensity of AlPcS₄ for all the samples measured. Among the five samples, Au NR-enhanced ¹O₂ generation in Au NR I (Fig. 3a), Au NR IV (Fig. 3d), and Au NR V (Fig. 3e) followed a relatively similar tendency, where the yield of ¹O₂ gradually increased first with the distance till reaching three bilayers (∼5.4 nm) and then decreased progressively with further bilayer incorporation. In contrast, Au NR II (Fig. 3b) and Au NR III (Fig. 3c) exhibited different behavior for the ¹O₂ generation. As shown in Fig. 3b and c, the lowest ¹O₂ yield was observed when AlPcS₄ was directly conjugated with Au NR. However, the signal intensity in Fig. 3c did not increase instantly when the polyelectrolyte bilayer spacer was introduced. This observation suggests that there is likely a quenching zone for Au NR II and Au NR III up to three bilayer distance (∼5.4 nm). The highest enhancement was observed with four bilayer spacer (∼7.4 nm). Note that LSPRs of Au NR II (Fig. 3b) and Au NR III (Fig. 3c) substantially overlap with emission band of AlPcS₄ shown in Fig. 3. As compared to Au NR, 50 nm SiO₂ nanoparticles without absorption band from visible light to NIR was chosen as control group. As shown in Fig. 3f, SiO₂-based complexes did not exhibit any correlation with distance for both AlPcS₄ fluorescence intensity and ¹O₂ generation, suggesting the significant effect of LSPR in Au NR-enhanced ¹O₂ generation and Au NR-enhanced fluorescence.

To gain insights into the phenomena in the context of our experimental results, we performed theoretical modeling of Au NR-enhanced ¹O₂ generation and Au NR-enhanced fluorescence. In accordance to the energy level diagram of typical PS, which is termed Jablonski diagram (see Fig. S1, ESI†), we depict relevant radiative and non-radiative processes and give the expressions for PS quantum yields. Intrinsically, Γ represents the rate constants for the various radiative (solid lines in Fig. S1†) and non-radiative (dashed lines in Fig. S1†) processes. PS excitation is generally achieved via one-photon transition between ground state (S₀) and the singlet excited state (S_n). Then, S_n relaxes to the lowest excited singlet excited state (S₁) via internal conversion (IC). For a ¹O₂ sensitizer, this is followed by efficient intersystem crossing (ISC) to the lowest triplet state (T₁).³² This latter process competes with radiative and non-radiative decays from the S₁ state. The lifetimes of the S₁ state are in the nanosecond range^33–35 that is too short to allow for significant interactions with the surrounding molecules. The lifetime of the T₁ state is in the micro- to milli-second range.^36,37 It is usually long enough for ground-state molecular oxygen (³O₂) to approach the sensitizer such that the excited-state ¹O₂ can be created via collision-dependent energy transfer.³⁸

For the AlPcS₄–Au NR system, the effects of a nearby Au NR on the relevant transitions rates of AlPcS₄, i.e., absorption rate (Γ_Abs), radiative decay rate (Γ_Rad), and non-radiative decay rate (Γ_NRad), will be introduced through near field intensity enhancement factor (M), radiative decay factor (M_Rad), and non-radiative decay factor (M_NRad), respectively. The reference quantities for free AlPcS₄ molecule (obtained with M = 1, M_Rad = 1 and M_NRad = 0) are denoted by the superscript 0.

The AlPcS₄–Au NR system is irradiated in such way that LSPR is induced in the Au NR, which gives rise to an increased electric field near the Au NR surface. The enhanced electric field will increase the amount of photons absorbed by the AlPcS₄ molecule. The enhancement of the near field intensity (|E|²) at the position of AlPcS₄ molecule is defined as

M = (|E|/|E₀|)² (1)

where |E₀|² is the incident intensity. The rate of photon absorption (Γ_Abs) is then increased by a factor M:^24,37

Γ_Abs = M × Γ⁰_Abs (2)

The excited AlPcS₄ molecule, described as an oscillating electrical dipole, is coupled at the emission wavelength to Au NR plasmon modes through near field dipole–dipole and, possibly, dipole–multipole interactions.³⁹ The presence of Au NR modifies Γ_Rad by a factor M_Rad written as

Γ_Rad = M_Rad × Γ⁰_Rad (3)

where Γ_Rad accounts for the energy reaching the far field. The coupling between Au NR also creates a new non-radiative decay channel associated with emission of a photon absorbed by Au NR and then Γ_NRad is given by

Γ_NRad = Γ⁰_NRad + M_NRad × Γ⁰_Rad (4)

where Γ⁰_NRad, non-radiative decay rate of free AlPcS₄, is only due to the process of IC.

Therefore, the fluorescence quantum yield (Q_Fluo) of AlPcS₄ at the presence of Au NR is given by

Q_Fluo = Γ_Rad/(Γ_Rad + Γ_NRad + Γ_ISC) (5)

where Γ_ISC is the rate of ISC.

The quantum yield of the T₁ state (defines the efficiency of ISC) is

Q_T = Γ_ISC/(Γ_Rad + Γ_NRad + Γ_ISC) (6)

The quantum yield of ¹O₂ generation (Q_Δ), i.e., the number of generated ¹O₂ molecules per the number of absorbed photons, can be expressed as

Q_Δ = Q_T × S_Δ (7)

where S_Δ is the fraction of T₁ molecules quenched by ³O₂ to produce ¹O₂.

The fluorescent rate (Γ_Fluo) of a single AlPcS₄ molecule can be expressed as the product of Γ_Abs and Q_Fluo. Finally, the fluorescence enhancement factor can be written as^24,40

EF_Fluo = Γ_Fluo/Γ⁰_Fluo = M × Q_Fluo/Q⁰_Fluo (8)

and more importantly, in an analogous way we can express the ¹O₂ enhancement factor as

EF_Δ = M × Q_T/Q⁰_T (9)

Based upon the experimental results, Au NR II and Au NR III exhibit different enhanced ¹O₂ generation and enhanced fluorescence behavior comparing to others. Therefore, Au NR II was chosen for further theoretical studies. To identify the LSPR of Au NR II, BEM was used to calculate the extinction spectrum of Au NR II, which consists of two components: scattering and absorption, whose relative contribution depends on size and shape of Au NR. The scattering component is known to be responsible for fluorescence enhancement (M_Rad) and the absorption component for fluorescence quenching (M_NRad). The Au NR cross-section spectra and field enhancement M were calculated in response to an incident plane wave. To evaluate the factors M_Rad and M_NRad, the plane wave excitation was replaced with a point electrical dipole adjacent to the Au NR. The simulated extinction and scattering cross-section spectra of the Au NR II are shown in Fig. 4a. The Au NR geometrical parameters and its orientation with respect to the propagation and polarization directions of the incident plane wave are also given in Fig. 4a. The peak of extinction spectrum is 700 nm (corresponding to the LSPR wavelengths), which is in a good agreement with Au NR II experimental LSPR wavelength (Table S2, ESI†). Because the extinction cross section is the sum of the absorption cross-section and scattering cross-section, it is evident from Fig. 4a that absorption strongly dominates over scattering.

Fig. 4 (a) Simulated extinction and scattering cross sections of Au NR II; (b) field intensity enhancement in the vicinity of Au NR II irradiated by a plane wave at the SPR wavelength 700 nm.

As given in Fig. 4a, in the illustrative case we present here, the excitation plane wave propagates in the z-direction and it is linear polarized in the x-direction or y-direction. The near field-enhanced polarization of the Au NR at the place of AlPcS₄ can be different from that of the incident plane wave and then it determines the direction of an induced electrical dipole in AlPcS₄. The excitation of the PS dipole was decomposed into the x-, y- and z-polarized components as they interact differently with Au NR plasmon modes. Therefore, the factors M, M_NRad and M_Rad were also decomposed and eqn (3)–(9) were correspondingly modified. Finally, using Einstein summation notation, the fluorescence or ¹O₂ enhancement factor (eqn (8) and (9)) can be expressed in the tensorial form:

EF_i = M_ij (field intensity enhancement factor) × Q_j (quantum yield enhancement factor) (10)

where i = x or y denotes the excitation plane wave polarization and j = x, y, z denotes the PS electrical dipole polarizations.

Fig. 4b shows a typical exponential decay of the near field intensity enhancement with the distance from the surface of the Au NR. It is also obvious that in the present of Au NR at the positions defined in Fig. 4b, the x-polarization of incident plane wave is converted into the z-polarization of near field with little coupling into the x-polarization of near field. On the other hand, the incident y-polarization results in the near field y-polarization. Remaining cross-polarization conversions are negligible. Therefore, in this case, the significant component of M are only M_xz and M_yy and, to a less extent, M_xx. These are responsible for the excitation of AlPcS₄ electrical dipole along the z-, y- and x-axis, respectively. For the sake of simplicity, the excitation light wavelength is also set to the LSPR peak of Au NR at 700 nm. The total decay factor M_Tot (equals to M_NRad + M_Rad) and M_Rad for the x-, y- and z-orientation (polarizations) of AlPcS₄ dipole are plotted in Fig. 5a as a function of distance from the Au NR surface. The factors M_Rad and M_NRad actually express the Γ_Rad and Γ_NRad normalized to Γ⁰_Rad (see eqn (3) and (4)). A relative orientation of AlPcS₄ molecule dipole with respect to the Au NR surface determines whether Γ_Rad is increased or decreased.⁴¹ For a tangentially oriented dipole, Γ_Rad is diminished because the molecular dipole and the dipole induced on the Au NR surface radiate out of phase. In contrast, Γ_Rad increases if the dipole is orientated perpendicularly toward the Au NR surface. Here, the fields of the PS molecular dipole and the induced dipole interfere constructively. Due to the simulated Au NR II geometry, the x-, y- and z-orientation (polarization) of the AlPcS₄ dipole result more likely in a mixture of the two cases. Nevertheless, it is obvious from Fig. 5a that the x-polarization is closest to the “tangential orientation” with the lowest M_Rad and the z-polarization to the “perpendicular orientation” with the largest M_Rad. Fig. 5a also indicates that the contribution of Γ_NRad to the enhanced total decay rate of AlPcS₄ molecule dominates over that of the Γ_Rad (correspondingly to the dominant absorption component of the extinction spectrum in Fig. 4a). As can be seen in Fig. 5a, the enhanced non-radiative decay through the resonant energy transfer between AlPcS₄ dipole and Au NR plasmon modes follows the orientation dependence similar to that of the radiative decay. The fluorescence and singlet oxygen enhancement factors are plotted against the distance in Fig. 5b. According to eqn (8) and (9), the enhancement factors (EF) are generally given by the product of the field enhancement (M) and respective quantum yield modification (Q/Q⁰). Due to a strong non-radiative energy transfer from the excited PS molecule into the nanoparticle metal (M_NRad ≫ 1) at the plasmon resonance wavelength, the quantum yields of both processes are quenched (Q/Q⁰ < 1). Thanks to the factor M_Rad, the quantum yield quenching is weaker for fluorescence and therefore it is overcome by the field enhancement (M > 1), contrary to the ¹O₂ generation enhancement. This correlation results in the fluorescence enhancement (83.5 fold at a distance of 1.5 nm for the x-incident polarization) while ¹O₂ generation is quenched over all simulated distances.

Fig. 5 (a) Distance dependence of the total decay factor M_Tot and radiative decay factor M_Rad for the different orientation of AlPcS₄ dipole with respect to the Au NR surface; (b) distance dependence of the fluorescence and ¹O₂ enhancement factors for the emission wavelength at 700 nm; (c) distance dependence of the fluorescence and ¹O₂ enhancement factors for the emission wavelength at 800 nm.

To reduce quenching of the triplet state quantum yield Q_T (see eqn (9)) through the enhanced total decay rate from the singlet state S₁, the PS emission wavelength was detuned from the LSPR wavelength and set to 800 nm for the test. Fig. 5c shows that the fluorescence enhancement reaches a somewhat lowered maximum (38.4 fold) at the distance of 3 nm and ¹O₂ production is now also enhanced with a flat maximum 5.4 fold around a distance of 11 nm (for the x-incident polarization). Considering this, later on, we conducted the calculations for the emission wavelength from 600 nm to 950 nm with the excitation wavelength maintained matching the LSPR wavelength at 700 nm and the distance set to 11 nm. It is evident from Fig. 6a that all radiative and non-radiative decay components, reducing the triplet state quantum yield, are maximized at the emission wavelength matching the LSPR wavelength and significantly diminished when the emission wavelength is detuned from the LSPR wavelength far enough. Fig. 6b indicates that the red shift of the emission wavelength provides a higher ¹O₂ enhancement, e.g., 9.4 fold at 900 nm (x-pol.). Fluorescence is still enhanced at the red-shifted wavelengths, 15 fold at 900 nm (x-pol.), which might be potentially utilized for the fluorescence imaging in the applications regarding PS delivery.

Fig. 6 (a) Wavelength dependence of the total decay factor M_Tot and radiative decay factor M_Rad for the different orientation of the PS dipole with respect to the Au NR surface; (b) wavelength dependence of the fluorescence and ¹O₂ enhancement factors.

We caution that when comparing the experimental and numerical data, one should be aware that the experimental results involve averaging over the excitation polarizations and AlPcS₄ molecule positions. We did not apply such averaging in the theoretical model in order to gain a fundamental insight into the constituent processes. Our numerical simulations were limited to the minimal PS–Au NR separation around 1 nm mainly with respect to BEM meshing requirements, nevertheless trends of the numerical data predict fast increase of non-radiative energy transfer and corresponding fluorescence and ¹O₂ quenching with further decreasing AlPcS₄–Au NR separation in accordance with our experimental results. In case of AlPcS₄ molecule adsorbed on the Au NR surface or at a distance of a few angstroms from Au NR, quantum effects might also be involved. The life time of singlet excited state S₁ and triplet state T₁ is nanosecond and microsecond, respectively, which makes it extremely challenge to experimentally track the process. We are not equipped for such experimental undertaking. Our comprehensive theoretical analysis has nevertheless provided important insights to the mechanism of energy transfer process.

Conclusions

In conclusion, we have shown that there is a significant interplay in Au NR-enhanced ¹O₂ generation and fluorescence intensity of AlPcS₄ with LSPR of Au NR and the distance between AlPcS₄ and Au NR as key parameters. With facile synthesis of Au NR of varying size and aspect ratio hence tunable LSPR and through easy incorporation of polyelectrolyte thin film spacers via LbL assembly, Au NR-enhanced ¹O₂ generation can be optimized by proper combination of Au NR of the preferred LSPR wavelength and the adequate spacing between AlPcS₄ and Au NR. Another interesting aspect is that Au NR-enhanced AlPcS₄ fluorescence presents different distance-dependent tendency compared to Au NR-enhanced ¹O₂ generation. Furthermore, the experimental results are supported by theoretically modeling, yielding significant insights into the fundamental mechanism in associated energy transfer processes and providing guidelines for the AlPcS₄–Au NR system optimization. Importantly, we have demonstrated that, in order to achieve the significant ¹O₂ generation enhancement, the system should be irradiated at the LSPR wavelength of Au NR to secure a strong excitation of AlPcS₄ placed at an optimal distance from Au NR. But at the same time, AlPcS₄ coupling with Au NR plasmon modes at the AlPcS₄ emission wavelength should be avoided by using AlPcS₄ with a larger red shift of the emission wavelength. This strategy could be applicable to other types of plasmonic nanostructures and PS. Therefore, we have developed a robust methodology to predict and manipulate the ¹O₂ generation, which will of great interest for all the ¹O₂-relevant applications, especially in PDT.

Acknowledgements

Y. H. thanks Stevens Institute of Technology for teaching assistantship support. This research was partly funded by the Czech Science Foundation (contract # P205/12/G118).

References

1. M. Triesscheijn, P. Baas, J. H. M. Schellens and F. A. Stewart, Oncologist, 2006, 11, 1034–1044.
2. M. Niedre, M. S. Patterson and B. C. Wilson, Photochem. Photobiol., 2002, 75, 382–391.
3. E. F. F. Silva, C. Serpa, J. M. Dabrowski, C. J. P. Monteiro, S. J. Formosinho, G. Stochel, K. Urbanska, S. Simões, M. M. Pereira and L. G. Arnaut, Chemistry, 2010, 16, 9273–9286.
4. Á. Juarranz, P. Jaén, F. Sanz-Rodríguez, J. Cuevas and S. González, Clin. Transl. Oncol., 2008, 10, 148–154.
5. C. A. Robertson, D. H. Evans and H. Abrahamse, J. Photochem. Photobiol., B, 2009, 96, 1–8.
6. B. C. Wilson and M. S. Patterson, Phys. Med. Biol., 2008, 53, R61–R109.
7. R. W. Redmond and I. E. Kochevar, Photochem. Photobiol., 2006, 82, 1178–1186.
8. Y. Zhang, K. Aslan, M. J. R. Previte and C. D. Geddes, J. Fluoresc., 2007, 17, 345–349.
9. M. K. Khaing Oo, Y. Yang, Y. Hu, M. Gomez, H. Du and H. Wang, ACS Nano, 2012, 6, 1939–1947.
10. Y. Yang, Y. Hu, H. Du and H. Wang, Chem. Commun., 2014, 50, 7287–7290.
11. Y. Yang, X. Yang, J. Zou, C. Jia, Y. Hu, H. Du and H. Wang, Lab Chip, 2015, 15, 735–744.
12. Y. Yang, N. Gao, Y. Hu, C. Jia, T. Chou, H. Du and H. Wang, Ther. Delivery, 2015, 6, 307–321.
13. Y. Hu, Y. Yang, H. Wang and H. Du, in SPIE BiOS, ed. W. J. Parak, M. Osinski and K. Yamamoto, International Society for Optics and Photonics, 2013, p. 85950B.
14. Y. Hu, Y. Yang, H. Wang and H. Du, ACS Nano, 2015, 9, 8744–8754.
15. B. Jang, J. Park, C. Tung, I. Kim and Y. Choi, ACS Nano, 2011, 5, 1086–1094.
16. Y. Hu, T. Chou, H. Wang and H. Du, J. Phys. Chem. C, 2014, 118, 16011–16018.
17. E. Kharlampieva, V. Kozlovskaya, J. Tyutina and S. A. Sukhishvili, Macromolecules, 2005, 38, 10523–10531.
18. U. Hohenester and A. Trügler, Comput. Phys. Commun., 2012, 183, 370–381.
19. A. J. MacRobert, S. G. Bown and D. Phillips, Ciba Found. Symp., 1989, 146, 4–12.
20. S. Dhami, J. J. Cosa, S. M. Bishop, M. S. C. Simpson and D. Phillips, in Proc. SPIE 2078, Photodynamic Therapy of Cancer, ed. G. Jori, J. Moan and W. M. Star, International Society for Optics and Photonics, 1994, pp. 475–482.
21. P. Johnson and R. Christy, Phys. Rev. B: Condens. Matter Mater. Phys., 1972, 6, 4370–4379.
22. K. Ray, R. Badugu and J. R. Lakowicz, Chem. Mater., 2007, 19, 5902–5909.
23. K. Itano, J. Choi and M. F. Rubner, Macromolecules, 2005, 38, 3450–3460.
24. R. Bardhan, N. K. Grady and N. J. Halas, Small, 2008, 4, 1716–1722.
25. Y. Zhang, K. Aslan, M. J. R. Previte and C. D. Geddes, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 1798–1802.
26. J. Aizpurua, P. Hanarp, D. S. Sutherland, M. Käll, G. W. Bryant and F. J. García de Abajo, Phys. Rev. Lett., 2003, 90, 57401.
27. L.-W. Nien, S.-C. Lin, B.-K. Chao, M.-J. Chen, J.-H. Li and C.-H. Hsueh, J. Phys. Chem. C, 2013, 117, 25004–25011.
28. C.-Y. Tsai, S.-P. Lu, J.-W. Lin and P.-T. Lee, Appl. Phys. Lett., 2011, 98, 153108.
29. Y. Zhang, K. Aslan, M. J. R. Previte and C. D. Geddes, Appl. Phys. Lett., 2007, 91, 23114.
30. R. Carminati, J. J. Greffet, C. Henkel and J. M. Vigoureux, Opt. Commun., 2006, 261, 368–375.
31. C. Flors, M. J. Fryer, J. Waring, B. Reeder, U. Bechtold, P. M. Mullineaux, S. Nonell, M. T. Wilson and N. R. Baker, J. Exp. Bot., 2006, 57, 1725–1734.
32. R. Toftegaard, J. Arnbjerg, H. Cong, H. Agheli, D. S. Sutherland and P. R. Ogilby, Pure Appl. Chem., 2011, 83, 885–898.
33. E. Zenkevich, E. Sagun, V. Knyukshto, A. Shulga, A. Mironov, O. Efremova, R. Bonnett, S. P. Songca and M. Kassem, J. Photochem. Photobiol., B, 1996, 33, 171–180.
34. M. Merchán, L. Serrano-Andrés, M. A. Robb and L. Blancafort, J. Am. Chem. Soc., 2005, 127, 1820–1825.
35. F. Wilkinson, W. P. Helman and A. B. Ross, J. Phys. Chem. Ref. Data, 1993, 22, 113.
36. W. Wu, H. Guo, W. Wu, S. Ji and J. Zhao, J. Org. Chem., 2011, 76, 7056–7064.
37. A. Ogunsipe, J.-Y. Chen and T. Nyokong, New J. Chem., 2004, 28, 822.
38. M. K. Kuimova, G. Yahioglu and P. R. Ogilby, J. Am. Chem. Soc., 2009, 131, 332–340.
39. A. Moroz, Opt. Commun., 2010, 283, 2277–2287.
40. P. Bharadwaj and L. Novotny, Opt. Express, 2007, 15, 14266.
41. T. Soller, M. Ringler, M. Wunderlich, T. A. Klar, J. Feldmann, H.-P. Josel, Y. Markert, A. Nichtl and K. Kürzinger, Nano Lett., 2007, 7, 1941–1946.

---

Footnotes

† Electronic supplementary information (ESI) available: Additional figures and tables as noted in the text. See DOI: 10.1039/c6ra22814c

‡ Present address: Department of Biological and Environmental Engineering, Cornell University, Ithaca, NY 14853, USA.

§ Present address: Department of Biomedical Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, Jiangsu 211106, China.

---